nomic behavior that are consistent with the earlier modeling of the determinants of parents' demand for children. The strong associations between variables measuring investments in children (withdrawing them from the work force and putting them in school), child death rates, and fertility suggest that in some countries behavioral self-regulating mechanisms may already be at work within the family to dampen today's rapid rates of population growth. Though statistical association does not establish causation, one interpretation of the available evidence would run as follows: The development process has somehow precipitated a sharp reduction in death rates, and subsequent changes in the structure and growth of demand for the factors of production have raised the relative returns to human capital. Specialization in both urban and rural markets has exerted pressures on crafts and home industries, depreciating the value of child labor and providing women with stronger incentives to find employment outside of the home. By raising child costs these aspects of development have induced parents to seek smaller families and to concentrate childbearing during a shorter span of years. The parents who succeed in averting unwanted additional births after middle age appear to invest increasing family resources in their children's education.
THE CONSEQUENCES OF POPULATION GROWTH
The current pace of population growth in low income countries is an unprecedented and recent phenomenon, and for this reason, if no other, our understanding of its social and economic significance for human welfare is limited. Previous attempts to deal with this subject are demonstrably inadequate. Yet some inferences for probable social consequences and their time dimensions can be drawn from investigations of the micro repercussions of contemporary population growth.
Two schemes for evaluating the economic consequences of population growth are widely used: a micro-analysis of births as human capital, and a macro-analysis of per capita economic growth in which changes in fertility affect the age distribution of the population. Neither is wholly satisfactory for deriving policy guidelines or for better understanding the basic problem.
The first approach to evaluating the gains of reducing fertility and population growth is associated with the work of Enke.* Applying the human capital analytic framework to a child, Enke shows that a child treated as a producer good generally yields a low rate of return, for the childhood period of dependency absorbs sufficient resources to offset much of the child's expected future earnings as a mature worker. Ohlin has shown that this result is not unique to low income economies. Indeed, under the highly favorable conditions of a high income economy experiencing rapid growth in labor
*For example, see (41, 42). the parents' environment and their reproductive and eco-